pybcsv 1.5.11.dev0__tar.gz → 1.5.14.dev0__tar.gz

pybcsv-1.5.14.dev0/CMakeLists.txt

@@ -0,0 +1,220 @@
+cmake_minimum_required(VERSION 3.15...3.30)
+project(pybcsv LANGUAGES C CXX)
+
+# ── Version from VERSION.txt ────────────────────────────────────────────
+# VERSION.txt is the single source of truth, mirroring cmake/GetGitVersion.cmake.
+# Deriving this from `git describe` instead meant a wheel built where git was
+# unavailable silently stamped whatever the last committed fallback happened to
+# be. sync_headers.py copies VERSION.txt next to this file for sdist builds, so
+# both the full checkout and the sdist find it.
+set(_PY_VERSION_FILE "${CMAKE_CURRENT_SOURCE_DIR}/../VERSION.txt")
+if(NOT EXISTS "${_PY_VERSION_FILE}")
+    set(_PY_VERSION_FILE "${CMAKE_CURRENT_SOURCE_DIR}/VERSION.txt")
+endif()
+if(NOT EXISTS "${_PY_VERSION_FILE}")
+    message(FATAL_ERROR
+        "[pybcsv] VERSION.txt not found in either ${CMAKE_CURRENT_SOURCE_DIR}/.. or "
+        "${CMAKE_CURRENT_SOURCE_DIR}. Run sync_headers.py or build from the project root.")
+endif()
+
+file(READ "${_PY_VERSION_FILE}" _PY_FILE_VERSION)
+string(STRIP "${_PY_FILE_VERSION}" _PY_FILE_VERSION)
+string(REGEX MATCH "^([0-9]+)\\.([0-9]+)\\.([0-9]+)$" _PY_FV_MATCH "${_PY_FILE_VERSION}")
+if(NOT _PY_FV_MATCH)
+    message(FATAL_ERROR
+        "[pybcsv] ${_PY_VERSION_FILE} contains '${_PY_FILE_VERSION}', not a valid X.Y.Z version.")
+endif()
+
+set(BCSV_VERSION_MAJOR ${CMAKE_MATCH_1})
+set(BCSV_VERSION_MINOR ${CMAKE_MATCH_2})
+set(BCSV_VERSION_PATCH ${CMAKE_MATCH_3})
+set(BCSV_VERSION_STRING "${_PY_FILE_VERSION}")
+
+message(STATUS "[pybcsv] Version: ${BCSV_VERSION_STRING}")
+
+# Set variables expected by cmake/version.h.in template
+set(VERSION_MAJOR ${BCSV_VERSION_MAJOR})
+set(VERSION_MINOR ${BCSV_VERSION_MINOR})
+set(VERSION_PATCH ${BCSV_VERSION_PATCH})
+set(VERSION_STRING ${BCSV_VERSION_STRING})
+
+# Generate version_generated.h into build dir
+set(VERSION_TEMPLATE "${CMAKE_CURRENT_SOURCE_DIR}/../cmake/version.h.in")
+if(EXISTS "${VERSION_TEMPLATE}")
+    configure_file("${VERSION_TEMPLATE}"
+        "${CMAKE_CURRENT_BINARY_DIR}/include/bcsv/version_generated.h" @ONLY)
+else()
+    # Fallback for standalone / sdist builds where ../cmake/ doesn't exist
+    file(WRITE "${CMAKE_CURRENT_BINARY_DIR}/include/bcsv/version_generated.h"
+        "// Auto-generated version header\n"
+        "#ifndef BCSV_VERSION_GENERATED_H\n"
+        "#define BCSV_VERSION_GENERATED_H\n"
+        "namespace bcsv { namespace version {\n"
+        "constexpr int MAJOR = ${BCSV_VERSION_MAJOR};\n"
+        "constexpr int MINOR = ${BCSV_VERSION_MINOR};\n"
+        "constexpr int PATCH = ${BCSV_VERSION_PATCH};\n"
+        "constexpr const char* STRING = \"${BCSV_VERSION_STRING}\";\n"
+        "} }\n"
+        "#endif\n"
+    )
+endif()
+
+# ── Find nanobind ───────────────────────────────────────────────────────
+find_package(Python 3.11 COMPONENTS Interpreter Development.Module REQUIRED)
+
+# Detect nanobind CMake dir via Python
+execute_process(
+    COMMAND ${Python_EXECUTABLE} -m nanobind --cmake_dir
+    OUTPUT_VARIABLE NB_DIR
+    OUTPUT_STRIP_TRAILING_WHITESPACE
+    COMMAND_ERROR_IS_FATAL ANY
+)
+list(APPEND CMAKE_PREFIX_PATH "${NB_DIR}")
+find_package(nanobind CONFIG REQUIRED)
+
+# ── Source files ────────────────────────────────────────────────────────
+# Prefer parent project's include/ (always available in full checkout / CI).
+# Fall back to local python/include/ (populated by sync_headers.py for standalone builds).
+set(PARENT_INCLUDE_DIR "${CMAKE_CURRENT_SOURCE_DIR}/../include")
+set(LOCAL_INCLUDE_DIR  "${CMAKE_CURRENT_SOURCE_DIR}/include")
+
+if(EXISTS "${PARENT_INCLUDE_DIR}/bcsv/bcsv.h")
+    set(BCSV_INCLUDE_DIR "${PARENT_INCLUDE_DIR}")
+    message(STATUS "[pybcsv] Using parent include: ${BCSV_INCLUDE_DIR}")
+elseif(EXISTS "${LOCAL_INCLUDE_DIR}/bcsv/bcsv.h")
+    set(BCSV_INCLUDE_DIR "${LOCAL_INCLUDE_DIR}")
+    message(STATUS "[pybcsv] Using local include: ${BCSV_INCLUDE_DIR}")
+else()
+    message(FATAL_ERROR "[pybcsv] Cannot find bcsv headers. Run sync_headers.py or build from the project root.")
+endif()
+
+set(LZ4_DIR "${BCSV_INCLUDE_DIR}/lz4-1.10.0")
+
+set(BCSV_SOURCES
+    pybcsv/bindings.cpp
+)
+
+set(LZ4_SOURCES
+    ${LZ4_DIR}/lz4.c
+    ${LZ4_DIR}/lz4file.c
+    ${LZ4_DIR}/lz4frame.c
+    ${LZ4_DIR}/lz4hc.c
+    ${LZ4_DIR}/xxhash.c
+)
+
+# ── Build the extension module ──────────────────────────────────────────
+nanobind_add_module(_bcsv STABLE_ABI ${BCSV_SOURCES} ${LZ4_SOURCES})
+
+target_include_directories(_bcsv PRIVATE
+    ${CMAKE_CURRENT_BINARY_DIR}/include
+    ${BCSV_INCLUDE_DIR}
+    ${LZ4_DIR}
+)
+
+set_target_properties(_bcsv PROPERTIES
+    CXX_STANDARD 20
+    CXX_STANDARD_REQUIRED ON
+    C_STANDARD 11
+)
+
+# ── Platform-specific flags ─────────────────────────────────────────────
+if(MSVC)
+    target_compile_options(_bcsv PRIVATE
+        /O2
+        /std:c++20
+    )
+elseif(APPLE)
+    # Apple Clang ships LLVM 18-era libc++ where jthread/stop_token are
+    # still experimental (guarded by _LIBCPP_HAS_NO_EXPERIMENTAL_STOP_TOKEN).
+    # The -fexperimental-library flag enables these features.
+    target_compile_options(_bcsv PRIVATE
+        -O3 -Wall -Wextra
+        -fexperimental-library
+        -mmacosx-version-min=13.3
+    )
+else()
+    # Linux: full C++20 with pthreads
+    target_compile_options(_bcsv PRIVATE
+        -O3 -Wall -Wextra
+    )
+    target_link_libraries(_bcsv PRIVATE pthread)
+endif()
+
+target_compile_definitions(_bcsv PRIVATE BCSV_HAS_BATCH_CODEC=1)
+
+# ── Install ─────────────────────────────────────────────────────────────
+install(TARGETS _bcsv LIBRARY DESTINATION pybcsv)
+
+# ── Bundled CLI tools ───────────────────────────────────────────────────
+# A curated subset of the BCSV CLI tools is compiled into the wheel and
+# installed into the environment's scripts dir (on PATH in every venv), so
+# `pip install pybcsv` ships the high-performance converters with a version
+# that always matches the bundled library. Sources come from the parent
+# checkout (CI / full clone) or from the copies sync_headers.py places under
+# python/src/ for standalone sdist builds.
+option(PYBCSV_BUILD_TOOLS "Build and bundle the BCSV CLI tools" ON)
+
+set(PARENT_TOOLS_DIR  "${CMAKE_CURRENT_SOURCE_DIR}/../src/tools")
+set(LOCAL_TOOLS_DIR   "${CMAKE_CURRENT_SOURCE_DIR}/src/tools")
+set(PARENT_SHARED_DIR "${CMAKE_CURRENT_SOURCE_DIR}/../src/shared")
+set(LOCAL_SHARED_DIR  "${CMAKE_CURRENT_SOURCE_DIR}/src/shared")
+
+if(EXISTS "${PARENT_TOOLS_DIR}/csv2bcsv.cpp")
+    set(BCSV_TOOLS_DIR  "${PARENT_TOOLS_DIR}")
+    set(BCSV_SHARED_DIR "${PARENT_SHARED_DIR}")
+elseif(EXISTS "${LOCAL_TOOLS_DIR}/csv2bcsv.cpp")
+    set(BCSV_TOOLS_DIR  "${LOCAL_TOOLS_DIR}")
+    set(BCSV_SHARED_DIR "${LOCAL_SHARED_DIR}")
+else()
+    set(BCSV_TOOLS_DIR "")
+endif()
+
+set(CLI11_INCLUDE_DIR "${BCSV_INCLUDE_DIR}/CLI11-2.6.2")
+
+if(PYBCSV_BUILD_TOOLS AND BCSV_TOOLS_DIR AND EXISTS "${CLI11_INCLUDE_DIR}/CLI11.hpp")
+    # Keep in sync with pybcsv/tools.py::TOOLS.
+    set(PYBCSV_TOOLS
+        csv2bcsv bcsv2csv bcsvHeader bcsvCast bcsvValidate
+        bcsvHead bcsvTail bcsvSampler bcsvCompare bcsvRepair bcsvGenerator)
+
+    # Compile the vendored LZ4 once, shared by every tool (the extension
+    # module compiles its own copy because nanobind owns that target).
+    add_library(pybcsv_tools_lz4 STATIC ${LZ4_SOURCES})
+    target_include_directories(pybcsv_tools_lz4 PUBLIC ${LZ4_DIR})
+
+    foreach(tool IN LISTS PYBCSV_TOOLS)
+        add_executable(${tool} "${BCSV_TOOLS_DIR}/${tool}.cpp")
+        target_include_directories(${tool} PRIVATE
+            ${CMAKE_CURRENT_BINARY_DIR}/include
+            ${BCSV_INCLUDE_DIR}
+            ${BCSV_TOOLS_DIR}
+            ${BCSV_SHARED_DIR}
+        )
+        target_include_directories(${tool} SYSTEM PRIVATE ${CLI11_INCLUDE_DIR})
+        target_link_libraries(${tool} PRIVATE pybcsv_tools_lz4)
+        target_compile_definitions(${tool} PRIVATE BCSV_HAS_BATCH_CODEC=1)
+        set_target_properties(${tool} PROPERTIES
+            CXX_STANDARD 20
+            CXX_STANDARD_REQUIRED ON
+        )
+        if(MSVC)
+            target_compile_options(${tool} PRIVATE /O2 /std:c++20)
+        elseif(APPLE)
+            target_compile_options(${tool} PRIVATE
+                -O3 -fexperimental-library -mmacosx-version-min=13.3)
+        else()
+            target_compile_options(${tool} PRIVATE -O3)
+            target_link_libraries(${tool} PRIVATE pthread)
+            target_link_options(${tool} PRIVATE -s)   # strip: ~60% smaller wheels
+        endif()
+    endforeach()
+
+    if(DEFINED SKBUILD_SCRIPTS_DIR)
+        install(TARGETS ${PYBCSV_TOOLS} RUNTIME DESTINATION "${SKBUILD_SCRIPTS_DIR}")
+    else()
+        install(TARGETS ${PYBCSV_TOOLS} RUNTIME DESTINATION bin)
+    endif()
+    message(STATUS "[pybcsv] Bundling CLI tools: ${PYBCSV_TOOLS}")
+else()
+    message(STATUS "[pybcsv] CLI tools not bundled (sources or CLI11 not found, or disabled)")
+endif()

{pybcsv-1.5.11.dev0 → pybcsv-1.5.14.dev0}/MANIFEST.in

@@ -17,4 +17,8 @@ prune tests
 prune examples
 global-exclude *.pyc
 global-exclude __pycache__
-global-exclude .DS_Store
+global-exclude .DS_Store
+# Bundled CLI tool sources (synced by sync_headers.py; compiled into the wheel)
+recursive-include src *.cpp
+recursive-include src *.h
+recursive-include src *.hpp

{pybcsv-1.5.11.dev0 → pybcsv-1.5.14.dev0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pybcsv
-Version: 1.5.11.dev0
+Version: 1.5.14.dev0
 Summary: High-performance Python bindings for the BCSV (Binary CSV) library with pandas integration
 Keywords: csv,binary,data-science,compression,performance,pandas,numpy,big-data,data-processing
 Author-Email: Tobias Weber <weber.tobias.md@gmail.com>
@@ -156,6 +156,40 @@ with pybcsv.ReaderDirectAccess() as da:
     print(da.read(100))  # alternative syntax
 ```
 
+## Bundled Native CLI Tools
+
+The wheel ships the core BCSV command-line tools as native, version-matched
+binaries — installed into the environment's scripts directory, so they are on
+PATH in an activated venv with no compiler or CMake required:
+
+| Tool | Purpose |
+|------|---------|
+| `csv2bcsv` | CSV → BCSV with validated type inference (see the main CLI docs) |
+| `bcsv2csv` | BCSV → CSV with row/column selection |
+| `bcsvHeader` | Show a file's schema |
+| `bcsvHead` / `bcsvTail` | Show the first/last N rows |
+| `bcsvCast` | Change/narrow column types |
+| `bcsvSampler` | Filter/project rows with a condition expression |
+| `bcsvValidate` | Integrity-check a file |
+| `bcsvRepair` | Recover data from truncated/corrupted files |
+| `bcsvCompare` | Compare two files (with float tolerance) |
+| `bcsvGenerator` | Generate synthetic test datasets |
+
+```bash
+csv2bcsv data.csv data.bcsv        # from an activated environment
+```
+
+```python
+import pybcsv
+
+pybcsv.tools.run("csv2bcsv", "data.csv", "data.bcsv", "--overwrite")
+print(pybcsv.tools.run("bcsvHeader", "data.bcsv").stdout)
+pybcsv.tools.path("csv2bcsv")      # absolute path, e.g. for custom pipelines
+```
+
+Source builds can opt out with `-DPYBCSV_BUILD_TOOLS=OFF`; the full tool suite
+(and docs) live in the main repository under `src/tools/`.
+
 ## Parquet Conversion Tools (CLI)
 
 Installing `pybcsv` provides two streaming command-line converters (require the

{pybcsv-1.5.11.dev0 → pybcsv-1.5.14.dev0}/README.md

@@ -110,6 +110,40 @@ with pybcsv.ReaderDirectAccess() as da:
     print(da.read(100))  # alternative syntax
 ```
 
+## Bundled Native CLI Tools
+
+The wheel ships the core BCSV command-line tools as native, version-matched
+binaries — installed into the environment's scripts directory, so they are on
+PATH in an activated venv with no compiler or CMake required:
+
+| Tool | Purpose |
+|------|---------|
+| `csv2bcsv` | CSV → BCSV with validated type inference (see the main CLI docs) |
+| `bcsv2csv` | BCSV → CSV with row/column selection |
+| `bcsvHeader` | Show a file's schema |
+| `bcsvHead` / `bcsvTail` | Show the first/last N rows |
+| `bcsvCast` | Change/narrow column types |
+| `bcsvSampler` | Filter/project rows with a condition expression |
+| `bcsvValidate` | Integrity-check a file |
+| `bcsvRepair` | Recover data from truncated/corrupted files |
+| `bcsvCompare` | Compare two files (with float tolerance) |
+| `bcsvGenerator` | Generate synthetic test datasets |
+
+```bash
+csv2bcsv data.csv data.bcsv        # from an activated environment
+```
+
+```python
+import pybcsv
+
+pybcsv.tools.run("csv2bcsv", "data.csv", "data.bcsv", "--overwrite")
+print(pybcsv.tools.run("bcsvHeader", "data.bcsv").stdout)
+pybcsv.tools.path("csv2bcsv")      # absolute path, e.g. for custom pipelines
+```
+
+Source builds can opt out with `-DPYBCSV_BUILD_TOOLS=OFF`; the full tool suite
+(and docs) live in the main repository under `src/tools/`.
+
 ## Parquet Conversion Tools (CLI)
 
 Installing `pybcsv` provides two streaming command-line converters (require the

pybcsv-1.5.14.dev0/VERSION.txt

@@ -0,0 +1 @@
+1.5.13